Revolutionizing Robotics: Penn Engineers Develop AI-Powered System to See Around Corners with Radio Waves
Imagine a robot navigating a warehouse, effortlessly seeing around corners to avoid obstacles and ensure safe passage. This groundbreaking technology, developed by engineers at the University of Pennsylvania, could soon become a reality. The team has created a system called HoloRadar that enables robots to perceive their surroundings beyond direct line of sight, using radio waves and advanced AI processing.
HoloRadar's unique approach involves turning walls into mirrors. Unlike visible light, radio waves have longer wavelengths, which can reflect off surfaces in predictable ways. This property allows the system to reconstruct three-dimensional scenes that are otherwise hidden from view. By capturing these reflections, HoloRadar can create a detailed picture of the environment, even in low-light conditions or darkness.
The system's design is tailored for real-world operations, ensuring mobility and real-time performance. It complements existing sensors in autonomous vehicles, such as LiDAR, by providing additional perception. This extra layer of insight gives machines more time to react to potential hazards, enhancing safety and performance.
At the core of HoloRadar's success is a custom AI system that combines machine learning with physics-based modeling. It enhances the resolution of raw radio signals, identifies multiple reflection paths, and uses a physics-guided model to trace reflections backward, reconstructing the actual 3D scene. This process is akin to walking into a room full of mirrors, where the AI must discern the true locations of objects.
The team has already tested HoloRadar in real indoor environments, successfully reconstructing walls, corridors, and hidden human subjects. Future research will focus on outdoor scenarios, presenting additional challenges due to longer distances and dynamic conditions. The ultimate goal is to enable machines to operate safely and intelligently in complex environments, mirroring human navigation skills.
This innovative research was conducted in the WAVES Lab at the University of Pennsylvania School of Engineering and Applied Science, with support from the University of Pennsylvania. The development of HoloRadar represents a significant step forward in robotics, promising to enhance the capabilities and safety of autonomous systems in various settings.
